100 GHz passively mode-locked Er:Yb:glass laser at 15 μm with 16-ps pulses

100 GHz passively mode-locked Er:Yb:glass laser at 1.5 µm with 1.6-ps pulses

Andreas E. H. Oehler,1,2,* Thomas Südmeyer,1 Kurt J. Weingarten,2 and Ursula Keller1

1Institute of Quantum Electronics, Department of Physics, ETH Zurich,Wolfgang-Pauli-Strasse 16, 8093 Zürich, Switzerland

2Time-Bandwidth Products AG, Technoparkstrasse 1, 8005 Zürich, Switzerland *Corresponding author: [email protected]

Abstract: We demonstrate a compact diode-pumped fundamentally modelocked Er:Yb:glass laser with a record high repetition rate of 101 GHz, generating 35 mW average power in 1.6-ps pulses in the 1.5-µm telecom window. This performance makes the laser an attractive pulse generator to be used in advanced high-speed return-to-zero (RZ) telecom systems.

©2008 Optical Society of America

OCIS codes: (320.7090) Ultrafast lasers; (140.4050) Mode-locked lasers; (140.3580) Lasers, solid-state

References and links

1. E. Yoshida, N. Shimizu, and M. Nakazawa, "A 40-GHz 0.9-ps regeneratively mode-locked fiber laser with a tuning range of 1530-1560 nm," IEEE Photon. Technol. Lett. 11, 1587-1589 (1999).

2. K. Sato, "Semiconductor light sources for 40-Gb/s transmission systems," IEEE J. Lightwave Technol. 20, 2035-2043 (2002).

3. L. Krainer, R. Paschotta, G. J. Spühler, I. Klimov, C. Y. Teisset, K. J. Weingarten, and U. Keller, "Tunable picosecond pulse-generating laser with a repetition rate exceeding 10 GHz," Electron. Lett. 38, 225-227 (2002).

4. U. Keller and A. C. Tropper, "Passively modelocked surface-emitting semiconductor lasers," Phys. Rep. 429, 67-120 (2006).

5. H. Lindberg, M. Sadeghi, M. Westlund, S. Wang, A. Larsson, M. Strassner, and S. Marcinkevicius, "Mode locking a 1550 nm semiconductor disk laser by using a GaInNAs saturable absorber," Opt. Lett. 30, 2793-2795 (2005).

6. D. Lorenser, D. J. H. C. Maas, H. J. Unold, A.-R. Bellancourt, B. Rudin, E. Gini, D. Ebling, and U. Keller, "50-GHz passively mode-locked surface-emitting semiconductor laser with 100 mW average output power," IEEE J. Quantum Electron. 42, 838-847 (2006).

7. A. E. H. Oehler, S. C. Zeller, K. J. Weingarten, and U. Keller, "Broad multiwavelength source with 50 GHz channel spacing for wavelength division multiplexing applications in the telecom C band," Opt. Lett. 33, 2158-2160 (2008).

8. A. Schlatter, B. Rudin, S. C. Zeller, R. Paschotta, G. J. Spühler, L. Krainer, N. Haverkamp, H. R. Telle, and U. Keller, "Nearly quantum-noise-limited timing jitter from miniature Er:Yb:glass lasers," Opt. Lett. 30, 1536-1538 (2005).

9. L. Krainer, R. Paschotta, S. Lecomte, M. Moser, K. J. Weingarten, and U. Keller, "Compact Nd:YVO4 lasers with pulse repetition rates up to 160 GHz," IEEE J. Quantum Electron. 38, 1331-1338 (2002).

10. C. Hönninger, R. Paschotta, F. Morier-Genoud, M. Moser, and U. Keller, "Q-switching stability limits of continuous-wave passive mode locking," J. Opt. Soc. Am. B 16, 46-56 (1999).

11. S. C. Zeller, T. Südmeyer, K. J. Weingarten, and U. Keller, "Passively modelocked 77 GHz Er:Yb:glass laser," Electron. Lett. 43, 32-33 (2007).

12. A. E. H. Oehler, S. C. Zeller, T. Südmeyer, U. Keller, and K. J. Weingarten, "Moving towards 100 GHz from a passively mode-locked Er:Yb:glass laser at 1.5 µm," in Conference on Lasers and Electro-Optics (Europe), (IEEE, 2007), talk CI6-1-THU.

13. U. Keller, K. J. Weingarten, F. X. Kärtner, D. Kopf, B. Braun, I. D. Jung, R. Fluck, C. Hönninger, N. Matuschek, and J. Aus der Au, "Semiconductor saturable absorber mirrors (SESAMs) for femtosecond to nanosecond pulse generation in solid-state lasers," IEEE J. Sel. Top. Quantum Electron. 2, 435-453 (1996).

14. U. Keller, "Recent developments in compact ultrafast lasers," Nature 424, 831-838 (2003). 15. G. J. Spühler, K. J. Weingarten, R. Grange, L. Krainer, M. Haiml, V. Liverini, M. Golling, S. Schon, and U.

Keller, "Semiconductor saturable absorber mirror structures with low saturation fluence," Appl. Phys. B 81, 27-32 (2005).

#104095 - $15.00 USD Received 14 Nov 2008; revised 16 Dec 2008; accepted 16 Dec 2008; published 17 Dec 2008

(C) 2008 OSA 22 December 2008 / Vol. 16, No. 26 / OPTICS EXPRESS 21930

16. J. Schröder, S. Coen, F. Vanholsbeek, and T. Sylvestre, "Passively mode-locked Raman fiber laser with 100 GHz repetition rate," Opt. Lett. 31, 3489-3491 (2006).

17. K. Sato, "100 GHz optical pulse generation using Fabry-Perot laser under continous wave operation," Electron. Lett. 37, 763-764 (2001).

18. U. Troppenz, J. Kreissl, W. Rehbein, C. Bornholdt, T. Gaertner, M. Radziunas, A. Glitzky, U. Bandelow, and M. Wolfrum, "40 GBit/s Directly Modulated InGaAsP Passive Feedback DFB Laser," in 32nd European Conference on Optical Communications (ECOC), (Cannes, 2006).

19. D. Kopf, G. Zhang, R. Fluck, M. Moser, and U. Keller, "All-in-one dispersion-compensating saturable absorber mirror for compact femtosecond laser sources," Opt. Lett. 21, 486-488 (1996).

20. J. Inoue, T. Isu, K. Akahane, N. Yamamoto, and M. Tsuchyida, "Characterization of highly stacked InAs quantum dot layers on InP substrate for a planar saturable absorber at 1.5 µm band," Phys. Stat. Solidi 3, 520-523 (2006).

21. D. J. H. C. Maas, A. R. Bellancourt, M. Hoffmann, B. Rudin, Y. Barbarin, M. Golling, T. Südmeyer, and U. Keller, "Growth parameter optimizationfor fast quantum dot SESAMs," Opt. Express 16, 18646-18656 (2008).

1. Introduction

Pulsed lasers with high repetition rate are important tools for advanced high-speed return-to-zero (RZ) data transmission systems, typically operating at 40 Gbit/s or higher. The developing 100 Gbit/s ethernet standard might also benefit from these lasers. Data streams are encoded on the pulse train with a modulator, which then only has to change its state between two successive pulses. This approach reduces the highly stringent demands on the slopes of the modulator, as the pulse shaping is already done by the laser source. These advantages provide a strong motivation to develop and improve laser sources such as harmonically mode-locked fiber lasers [1], hybrid mode-locked semiconductor lasers [2], and passively mode-locked solid-state lasers [3] that operate in this regime of high repetition rates. Another interesting technology is the SESAM-modelocked VECSEL (Vertical External Cavity Surface Emitting Laser) [4] which has achieved repetition rates up to 3 GHz in the 1.5 µm telecom window to date [5]. Repetition rates up to 50 GHz have been demonstrated [6], however at a wavelength of 960 nm, which is not suitable for telecommunication applications. Passively mode-locked Er:Yb:glass lasers represent a simple way to directly generate a high repetition rate pulse train in the telecom C-band. The pulse generation is passive, therefore no high frequency electronics is required. In addition we have previously demonstrated very stable frequency combs [7] and nearly quantum noise limited timing jitter [8]. In current state-of-the-art high bit-rate data transmission systems running at 100 Gbit/s, for example, the pulsed laser source usually runs at a much lower repetition rate and needs to be time multiplexed to the desired 100 GHz. Generating a 100-GHz pulse train with high average output power directly with a mode-locked laser greatly simplifies the system design, as the multiplexing device as well as the amplifiers can be omitted. Additionally, the pulse-to-pulse phase stability of these fundamentally mode-locked lasers allows for phase-sensitive modulation techniques such as difference phase-shift keying (DPSK).

Here we demonstrate a SESAM mode-locked diode-pumped Er:Yb:glass laser, which operates at a record high repetition rate of 101 GHz, generating 35 mW average power in 1.6 ps pulses. Previously, such repetition rates directly generated by a fundamentally mode-locked laser were only achieved around 1-µm wavelength where gain media with larger emission cross section are available [9], which strongly reduces the tendency for Q-switching instabilities [10]. Compared to previous results [11, 12] we have further increased the pulse repetition rate, reduced the pulse duration by a factor of two and increased the average output power by a factor of three.

2. Experimental setup

For 100 GHz, fundamental mode locking leads to an extremely small cavity length of about 1.26 mm (taking into account that the beam-path inside the gain medium, with a refractive index of 1.53 at 1550 nm, represents a substantial fraction of the overall cavity length) as the repetition rate of the laser is only determined by the roundtrip time of the single pulse



traveling inside the cavity. The resonator is folded into a V-shape, which leads to a very compact cavity geometry. Fig. 1 shows a schematic of the setup, Fig. 2 a photograph of the actual laser cavity.

Fig. 1. Schematic of experimental setup of the 101 GHz laser cavity. The Er:Yb:glass is flat on one side, coated with a 1.2% output coupler and under Brewster’s angle inside the cavity. The collimated pump beam is focused into the gain glass such that it is mode-matched to the laser mode. The output beam is collimated by the same lens. Both beams are separated afterwards by a dichroic mirror. SESAM: semiconductor saturable absorber mirror, OC: output coupler.

Fig. 2. Photograph of actual V-shaped cavity of a SESAM modelocked Er:Yb:glass laser. The cavity length is at this point mechanically limited because the curved mirror almost touches the Er:Yb:glass gain element.

The Er:Yb:glass laser gain medium is well-suited for telecom applications, because it can

be pumped with reliable standard 980-nm high brightness laser diodes, and its gain bandwidth covers the entire C-band. However, its small emission cross section typically limits the ability to operate at high repetition rates without Q-switched mode locking (QML) [10]. To overcome this challenge, we have optimized the semiconductor saturable absorber mirror (SESAM) [13] for low saturation fluence and moderate modulation depth, and the cavity design for small mode areas in the gain medium as well as on the SESAM. The improved

gain glass with OC coating

curved mirror

SESAM



cavity geometry has a gain element flat on one side under normal incidence, with a 1.2% output coupling coating and cut at Brewster’s angle on the other side. The folding angle of the curved folding mirror is calculated such that it compensates the astigmatism introduced by the gain element as it is shown in Fig. 3. The cavity has a round and stigmatic output beam, which can be efficiently coupled into a standard single-mode fiber.

30

25

20

15

10

outp

ut d

iver

genc

e [m

rad]

50403020folding angle [deg]

tan sag

Fig. 3. The laser cavity was designed such that the folding mirror compensates for the astigmatism introduced by the gain element. At a folding angle of 43 degrees the output beam is round and stigmatic.

At such high repetition rates, the intracavity pulse energies become very low (around 0.35

pJ in the described laser). To suppress Q-switched mode-locking a small mode radius of about 15 µm is used in the gain medium. In addition, a tight focus of about 4 µm radius on the SESAM is needed for a sufficient saturation of the absorber [14], therefore the folding mirror is strongly curved (radius of curvature of less than 1 mm). The SESAM consists of an InGaAs quantum well embedded in a low finesse design, resulting in low saturation fluence and a modulation depth below 1% [15].

3. Results

We have generated a pulse train at 101.5 GHz with pulses of 1.6 ps pulse duration and good extinction ratio (Fig. 4). Normally we would use a fast photodiode with a microwave spectrum analyzer to determine the repetition rate, however such devices operating to beyond 100 GHz were not available for this work. Therefore the repetition rate was determined with an autocorrelator from Femtochrome Research Inc. (model FR-103MN) by measuring the distance between the peaks of the autocorrelation and the first cross-correlations. The autocorrelator uses a rotating prism-pair as its varying delay arm, which is not internally compensated to provide a linear time scale at its analog output. If a large time-window is measured, some deviations from linearity have to be corrected. To calibrate the autocorrelator, we used a laser with 200-fs pulses at 1.5 µm wavelength to measure the temporal shift of the pulse peak position with respect to the length of the fixed delay arm. We could fit this dependence with a 2nd order polynomial to avoid an error of 10% over a scan length of 40 ps compared to a linear fit. As an additional verification, we also measured the distance between the autocorrelation- and the cross-correlation peaks by varying the fixed delay arm. We measured a path-length of 1.48 mm which corresponds to a pulse repetition rate of 101.28 GHz. This is in good agreement with the above result. The high average output power of 35 mW allowed us to directly measure the autocorrelation from the laser output without any further amplification, thus avoiding noise- and dispersion-issues which could be introduced by using a fiber-amplifier to amplify the pulse train. The optical spectrum shown in Fig. 5 is centered at 1534.8 nm with a full width at half maximum of



2.6 nm. This results in a time bandwidth product (TBP) of 0.53, which is 1.7 times the transform limit for sech2 pulses.

1.0

0.8

0.6

0.4

0.2

0.0

auto

corr

. sig

nal [

a.u.

]

-15 -10 -5 0 5 10 15time delay [ps]

TR

Fig. 4. Autocorrelation trace of 101 GHz pulse train including cross-correlations (solid line) with fit-curve using ideal sech2 pulses (grey dashed line). Cavity roundtrip time TR = 9.85 ps. The pulses have a duration of 1.6 ps.

1.0

0.8

0.6

0.4

0.2

0.0

spec

tral i

nt. [

a.u.

]

1544154015361532wavelength [nm]

FWHM: 2.6 nm

-60

-40

-20[d

B]

154015351530

Fig. 5. Optical spectrum taken with a resolution bandwidth of 0.01 nm (solid line). The full width at half maximum determined by using ideal sech2 fit (grey dashed line) is 2.6 nm centered at 1534.8 nm. The mode separation is 0.8 nm. The inset shows the spectrum on a logarithmic scale.

We measured the mode spacing in the optical spectrum to be 0.8 nm, which corresponds to a repetition rate of 101.81 GHz, also confirming the repetition rate as measured with the autocorrelator. The Q-switching threshold is below 2.8 mW output power. Above this power level, the pulse train is stable, which we could verify with the autocorrelator. Fig. 6 shows the laser slope above the QML threshold together with the optical-to-optical efficiency.

4. Conclusion and outlook

We have demonstrated a fundamentally SESAM mode-locked Er:Yb:glass laser with a repetition rate of 101 GHz at 1534.2 nm and an average output power of 35 mW in 1.6-ps pulses. The mode locking is self-starting and stable. To the best of our knowledge, this is the highest repetition rate generated directly from a passively mode-locked solid-state laser oscillator operating in the 1.5-µm telecom window. Compared to previous results, we have increased the repetition rate to exceed 100 GHz while increasing the average output power by



a factor of three and reducing the pulse duration by a factor of two. The high average output power at high repetition rate, the good pulse quality and pulse-to-pulse phase stability, and the compact and simple setup make this laser very competitive compared to harmonically mode-locked fiber lasers [16], semiconductor lasers [17] and distributed feedback (DFB) lasers [18], for high-speed data-transmission applications.

40

30

20

10

0

outp

ut p

ower

[mW

]

4003002001000pump power [mW]

0.10

0.08

0.06

0.04

0.02

0.00

optical efficiency

Fig. 6. Mode-locked optical output power of the 101 GHz Er:Yb:glass laser. We could obtain an average output power of 35 mW at a pump power of 370 mW. The QML-threshold was at 2.8 mW, above this power-level the mode-locking was stable. The optical-to-optical efficiency is close to 10%.

The 1.6-ps pulses are chirped to 1.7 times the transform limit, which means that the

output pulses could be compressed with external dispersion compensation into the sub-ps regime. Furthermore the laser cavity could be dispersion optimized to obtain sub-ps transform-limited pulses directly out of the laser, for example by optimizing the group delay dispersion of the dielectric coating on the folding mirror, or with a Gires-Tournois interferometer structure included in the SESAM design [19].

A fundamentally mode-locked laser with a substantially higher pulse repetition rate would require a different mechanical setup: in the present design the cavity-length can not be reduced much further as the individual components start to interfere. A monolithic cavity design with a straight resonator would be necessary to overcome these mechanical constraints. However, the SESAM saturation requires a very small mode size on the absorber which can only be obtained by strong focusing. This leads to a small Rayleigh-range of the cavity mode at the absorber preventing a small mode size inside the gain element which is needed to suppress Q-switched mode-locking. SESAMs with a substantially lower saturation fluence (e.g. quantum-dot SESAMs [20, 21]) and gain media with a higher gain cross-section could solve this issue.

Acknowledgments

This work was supported by the European Commission through CRAFT project MULTIWAVE (018074) and ETH Zurich with research grant TH-04 07-3 “Multiwave”.



100 GHz passively mode-locked Er:Yb:glass laser at 15 μm with 16-ps pulses

Documents

Transcript of 100 GHz passively mode-locked Er:Yb:glass laser at 15 μm with 16-ps pulses